Fig. 2: Visualization and analysis of Dense-UNet features for segmentation task before training the network.
From: Revealing hidden patterns in deep neural network feature space continuum via manifold learning
Here, B3-L2 denotes the 2nd layer of the 3rd dense block, B4-L8 denotes the 8th layer of the 4th dense block, B6-L2 denotes the 2nd layer of the 6th dense block, and B7-L8 denotes the last layer before the final output. t-SNE, UMAP and MDA results are shown in (a, b, c) respectively for training and testing datasets at different network layers. The colorbar denotes the normalized manifold distance. Source data are provided as a Source Data file.
